There should have been another control - when the odd man out is of an intermediate color with no well-known name for the participants that have the blue-green distinction, e.g. something like teal or turquoise. In this case I'd predict that the blue-green distinction would be a disadvantage.